INDEX

Explanations

terms related to authorship and attribution in written content

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

XMLSchema

-0.39

 Nagel

-0.39

]!=

-0.38

Nau

-0.37

cooked

-0.36

 гер

-0.35

dos

-0.35

 Apfel

-0.35

 Coates

-0.35

az

-0.35

POSITIVE LOGITS

ViewHolder

2.73

 ViewHolder

1.28

 viewHolder

0.98

viewHolder

0.81

enterOuterAlt

0.67

 onOptions

0.63

sidemargin

0.60

localctx

0.54

MigrationBuilder

0.52

StoryboardSegue

0.51

Activations Density 0.000%

terms related to authorship and attribution in written content

No Comments

No Known Activations